| // HDLbits example excercise "vector2" | |
| // https://hdlbits.01xz.net/wiki/Vector2 | |
| module endian_switch | |
| #( | |
| parameter BYTE_SIZE = 8, | |
| parameter INPUT_BYTES = 4 | |
| ) | |
| ( | |
| input [INPUT_BYTES*BYTE_SIZE-1:0] in, | |
| // both array and flat versions are driven | |
| // connect up as required | |
| output [INPUT_BYTES*BYTE_SIZE-1:0] out, | |
| output [INPUT_BYTES-1:0][BYTE_SIZE-1:0] out_array | |
| );// | |
| // $size gives number of bits in a vector | |
| localparam NUM_BYTES = INPUT_BYTES; | |
| // generate variable | |
| genvar i; | |
| generate | |
| for (i = 1;i<=NUM_BYTES;i++) begin : endian_switch_for // loop name | |
| // mimics the same format as intermediate, but this just automatically expands based on the constants | |
| assign out[i*BYTE_SIZE-1 -: BYTE_SIZE] = in[(NUM_BYTES-i+1)*BYTE_SIZE-1 -: BYTE_SIZE]; | |
| // Array driven too | |
| // for arrays, they are indexed in the same order as declared | |
| // out_array is declared as [INPUT_BYTES-1:0][BYTE_SIZE-1:0] | |
| // so assign out_array[x][y] | |
| // x -> INPUT_BYTES dimension | |
| // y -> BYTE_SIZE dimension. | |
| // right-side dimensions can be left off | |
| // assign out_array[x] is OK (as seen below), but there is no way to index [y] on its own | |
| // so be careful to get the order right according to the application. | |
| // else manual dimension switch is required | |
| assign out_array[i-1] = in[(NUM_BYTES-i+1)*BYTE_SIZE-1 -: BYTE_SIZE]; | |
| end | |
| endgenerate | |
| endmodule | |